INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

At above critical concentrations, essential transition metal ions such as Ni²⁺, Co²⁺, and Zn²⁺ are toxic, being, for example, potent inhibitors of processes such as respiration and photosynthesis (see, for example, references 4, 22, 28, and 39). In addition, transition metals are required for the catalytic activity of a number of enzymes because of their redox activity and their high charge density, which allows the polarization of substrates and the stabilization of transition state intermediates. Bacteria have evolved sensing, sequestering, and transport systems that allow a precise homeostasis for these metals.

During the last years it has became clear that microbial Ni²⁺, Zn²⁺, and Co²⁺ uptake is mediated by nonspecific transport systems for divalent cations (33) and by high-affinity specific systems. Two types of high-affinity transporters have been identified: (i) multicomponent ATP-binding cassette transport systems (such as NikABCDE for Ni²⁺ or ZnuABC for Zn²⁺) (31, 37) and (ii) one-component transporters (such as NixA, UreH, HupN, and HoxN for Ni²⁺ and NhlF for Co²⁺) (12, 16, 23, 27, 29) which are integral membrane proteins with eight transmembrane-spanning helices.

Most of the studies on Co²⁺, Zn²⁺, and Ni²⁺ export and resistance have been carried out with the soil chemolithotrophic Alcaligenes strains (now designed Ralstonia), where three sequence-related divalent cation efflux operons, called czc (for Cd²⁺, Zn²⁺, and Co²⁺ resistance) (32), cnr (for Co²⁺ and Ni²⁺ resistance) (25), and ncc (for Ni²⁺, Co²⁺, and Cd²⁺ resistance) (47), have been described. Zn²⁺-dependent efflux ATPases have been recently characterized for Escherichia coli (zntA) (3) and for the cyanobacterium Synechocystis sp. strain PCC 6803 (ziaA) (51). ZntA and ZiaA belong to the P-type ATPase family (recently reviewed in reference 40), which includes the bacterial Cd²⁺ transporter CadA (34) and bacterial Cu²⁺ transporters (20, 35, 38).

Much less is known about how Ni²⁺, Zn²⁺, and Co²⁺ are sensed and how metal binding provokes protein conformational changes that determine regulatory responses. Two types of Zn²⁺-responsive regulators have been recently described. ZntR is a MerR-like transcriptional activator of zntA expression in E. coli (7). In contrast, Synechococcus sp. strain PCC 7942 SmtB (19), Synechocystis sp. strain PCC 6803 ZiaR (51), and Staphylococcus aureus ZntR (49) are transcriptional repressors that belong to the ArsR-SmtA family of helix-turn-helix DNA binding proteins. Although the overall ternary structure for these repressors is conserved, the metal binding site may be unique for each specific member of the family (9). Regulation of the czc efflux operon of Ralstonia eutropha is currently under active study, and at least three proteins, CzcD, CzcR, and CzcS, seem to be involved in metal sensing (55). The only nickel-specific responsive regulator reported is NikR, a Fur-related DNA binding protein that represses the transcription of the E. coli nikABCDE operon in the presence of high Ni²⁺ concentrations (11). Finally, no Co²⁺-specific sensor proteins have been reported so far.

We report in the present work the identification and characterization of a metal-regulated gene cluster in the unicellular cyanobacterium Synechocystis sp. strain PCC 6803. The cluster comprises nine open reading frames (ORFs) organized into five transcriptional units and seems to be responsible for Ni²⁺, Co²⁺, and Zn²⁺ homeostasis in Synechocystis sp. strain PCC 6803.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Strains and growth conditions. Synechocystis sp. strain PCC 6803 was grown photoautotrophically at 30°C in BG11 medium (43) supplemented with 1 g of NaHCO₃ per liter (BG11C) and bubbled with a continuous stream of 1% (vol/vol) CO₂ in air under continuous illumination (50 µmol of photons per m² per s; white light from fluorescent lamps). For plate cultures, BG11C liquid medium was supplemented with 1% (wt/vol) agar. Kanamycin was added to a final concentration of 50 to 200 µg/ml when required. BG11C medium was supplemented with different concentrations of ZnSO₄, CdCl₂, CoCl₂, CuSO₄, NiSO₄, and MgCl₂ when indicated.

Insertional mutagenesis of Synechocystis sp. strain PCC 6803 genes. Loci slr0794, slr0796, slr0797, and sll0794 were inactivated by interruption with a kanamycin resistance cassette (C.K1) (14). For this, DNA fragments containing loci slr0794, slr0796, slr0797, and sll0794 were amplified by PCR from the cosmid CS1377 (provided by Kazusa DNA Research Institute) and cloned into pGEM-T (Promega). slr0794 was amplified by using oligonucleotides nit1 and nit2 (Table 1) and cloned into pGEM-T to generate pNIQ7. Targeting vectors were generated by inserting the C.K1 cassette into the EcoRI site of slr0794 in the same orientation as the nrs operon [pNIQ8(+)] or in the inverse orientation [pNIQ8()]. slr0796 was amplified by using oligonucleotides nrp1 and nrp2 (Table 1) and cloned into pGEM-T to generate pNIQ1. Targeting vectors were generated by inserting the C.K1 cassette into the BstEII site of slr0796 in the same orientation as the nrs operon [pNIQ2(+)] or in the inverse orientation [pNIQ2()]. slr0797 was amplified by using oligonucleotides cor1 and cor2 (Table 1) and cloned into pGEM-T to generate pNIQ10. The targeting vector was generated by inserting the C.K1 cassette into the EcoNI site of slr0797 in the opposite orientation to the slr0797 gene (pNIQ12). sll0794 was amplified by using oligonucleotides mr1 and mr2 (Table 1) and cloned into pGEM-T to generate pNIQ3. The targeting vector was generated by inserting the C.K1 cassette into the HindIII site of sll0794 in the same orientation as the sll0794 ORF [pNIQ4(+)]. All targeting vectors were used to transform Synechocystis sp. strain PCC 6803 strain as previously described (15).

RNA isolation and Northern blot hybridization. Total RNA was isolated from 25-ml samples of Synechocystis sp. strain PCC 6803 cultures at the mid-exponential growth phase (3 to 5 µg of chlorophyll/ml). Extractions were performed by vortexing cells in the presence of phenol-chloroform and acid-washed baked glass beads (0.25- to 0.3-mm diameter; Braun, Melsungen, Germany) as previously described (17).

Purification of GST-C-NrsD and metal affinity chromatography. The last 126 bp of the nrsD ORF, which encodes the last 42 amino acids of NrsD, was amplified by PCR using the oligonucleotides nrh1 and nrh2 (Table 1). The resulting DNA fragment was digested with EcoRI and XhoI and cloned into pGEX-4T-3 in phase with the GST gene to generate pGEX-C-NrsD. GST-C-NrsD fusion protein and GST were expressed in E. coli BL21 from the plasmids pGEX-C-NrsD and pGEX-4T-3, respectively. One liter of culture was grown in Luria broth medium supplemented with 2% glucose to an optical density at 600 nm of 0.6, induced with 1 mM isopropyl--D-thiogalactopyranoside for 2.5 h, harvested by centrifugation, and resuspended in 20 ml of phosphate-buffered saline buffer (150 mM NaCl, 16 mM Na₂HPO₄, 4 mM NaH₂PO₄, pH 7.2) supplemented with 1 mM phenylmethylsulfonyl fluoride. Cells were broken by sonication, and insoluble debris was pelleted by centrifugation at 18,000 × g for 15 min. The supernatant was then applied to a glutathione-agarose bead column (Pharmacia) (1-ml bed volume). After extensive washing with phosphate-buffered saline buffer, GST or GST-C-NrsD proteins were eluted with 3 ml of 50 mM Tris HCl (pH 8.0) containing 10 mM reduced glutathione. Glutathione was then removed by gel filtration in a Sephadex G-25 column. Interaction of GST-C-NrsD or GST with Ni²⁺, Co²⁺, Zn²⁺, Cu²⁺, or Mg²⁺ was investigated by metal ion affinity chromatography. A 0.5-ml portion of His-bind resin (Novagen) was loaded with 0.5 ml of 0.5 M ZnSO₄, CoCl₂, CuSO₄, NiSO₄, or MgCl₂ in water and then equilibrated in 0.5 M sodium chloride-50 mM Tris HCl (pH 8.0) (buffer A). About 30 µg of purified GST-C-NrsD or GST proteins were applied to the columns. Unbound proteins were removed by washing with buffer A. Bound polypeptides were eluted with 0.5 ml of 0.4 M imidazole in buffer A. Proteins were then analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (24) and Coomassie blue staining. Quantities of bound and unbound proteins were determined by the method of Bradford (5).

Computer methods. The BLAST program (1) was used to screen the translated nucleotides databases. The CLUSTAL X program was used to generate sequence alignments (53). Putative membrane-spanning regions were identify using different algorithms (18, 57).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

A metal-regulated gene cluster in Synechocystis sp. strain PCC 6803. Analysis of the fully sequenced Synechocystis sp. strain PCC 6803 genome (21) allowed us to identify a region of the chromosome containing three ORFs whose deduced amino acid sequences are clearly related to metal transport proteins (see below). The region extends 12 kb and comprises nine genes organized into five putative transcription units (Fig. 1A). ORF slr0798 has been reported to encode a Zn²⁺-dependent efflux ATPase (ZiaA) whose expression is Zn²⁺ dependent (51). A putative operon composed of two ORFs, sll0793 and sll0792, separated by 11 bp appears upstream from the ziaA gene and in the opposite orientation. sll0792 encodes ZiaR, the transcriptional repressor of ziaA. sll0793 is a putative membrane protein that does not share significant homology with any other protein in the EMBL-GenBank database (51). Metal-dependent expression of the remaining three putative transcriptional units was analyzed by Northern blotting. For this, four probes were used to hybridize total RNA obtained from mid-log-phase Synechocystis sp. strain PCC 6803 cells grown in BG11C medium and exposed during 1 h to a 15 µM concentration of either ZnSO₄, CdCl₂, CoCl₂, CuSO₄, NiSO₄, or MgCl₂ (Fig. 1B). Control cells were not exposed to added metals. Probes a (internal to slr0793) and b (internal to slr0796) hybridized strongly with RNA obtained from Ni²⁺-exposed Synechocystis sp. strain PCC 6803 cells and weakly with RNA from Co²⁺-exposed cells. Probe c (ORF sll0794) showed no hybridization with RNA from any of the conditions tested. Probe d, corresponding to ORF slr0797, hybridized strongly with RNA from Co²⁺-exposed cells and weakly with RNA from Zn²⁺-exposed cells. Transcript levels of the RNase P RNA (rnpB gene) (56) remained unchanged under all tested conditions (Fig. 1B).

View larger version (62K): [in this window] [in a new window]   FIG. 1.   Metal-dependent expression of the Synechocystis transition metal-resistant cluster. (A) ORF organization of the metal-regulated cluster from Synechocystis sp. strain PCC 6803. (B) Total RNA was isolated from mid-log-phase Synechocystis sp. strain PCC 6803 cells exposed for 1 h to a 15 µM concentration of the indicated metal ions. Control cells were not exposed to added metals (). Fifteen micrograms of total RNA was denatured, separated by electrophoresis in a 1.2% agarose gel, blotted, and hybridized with probes a to d as indicated in panel A (see Materials and Methods). The filters were stripped and rehybridized with an rnpB gene probe as a control. Estimated sizes of the transcripts (in nucleotides [n]) are indicated.

nrs is a nickel resistance operon. As shown above, a similar pattern of induction was found using probes from ORFs slr0793, and slr0796 (Fig. 1B). The fact that both probes hybridized with RNA of about 6 kb, together with the structure of the region, suggests that ORFs slr0793, slr0794, slr0795, and slr0796 form a transcriptional unit. Since Ni²⁺ provoked the highest induction of this transcription unit, the genes were named nrs for Ni²⁺ response system. The nrs induction dependence on concentration was studied by using Northern blot experiments. The operon was induced at Ni²⁺ concentrations of above 0.45 µM. An Ni²⁺ concentration of above 17 µM did not provoke higher accumulation of the nrs mRNA (Fig. 2A and data not shown). Time course analysis indicated that nrs mRNA was already induced 15 min after metal addition and increased almost linearly, at least during the first 4 h of treatment (Fig. 2B and C).

View larger version (30K): [in this window] [in a new window]   FIG. 2.   Ni2+ concentration dependence and time course of the expression of nrs. (A) The indicated concentration of NiSO4 was added to mid-log-phase Synechocystis sp. strain PCC 6803 cells grown in BG11C medium. After 1 h, cells were harvested and total RNA was isolated, processed, and hybridized as described for Fig. 1, using an nrsB gene probe (probe a [Fig. 1]). (B) A 17 µM concentration of NiSO4 was added to mid-log-phase Synechocystis sp. strain PCC 6803 cells grown in BG11C medium. Samples for total RNA isolation were taken at the indicated times. RNA was processed and hybridized as for Fig. 1, using an nrsB gene probe. (C) Radioactive signals of the time course experiment were quantified with a InstantImager Electronic Autoradiography apparatus. Levels of nrs operon mRNA were normalized with the rnpB signal, and plots of relative mRNA levels versus time were drawn.

View larger version (26K): [in this window] [in a new window]   FIG. 3.   Ni2+ tolerance of Synechocystis nrsA::C.K1 and nrsD::C.K1 mutants. (A) Schematic representation of the nrs genomic region in the wild-type strain and sites of insertion of the C.K1 cassette in the nrsA::C.K1 and nrsD::C.K1 mutants. The C.K1 cassette was inserted in both orientations, as indicated. B, BstEII; E, EcoRI. (B) Ni2+ tolerance of wild-type Synechocystis sp. strain PCC 6803 (WT) and Synechocystis nrsA::C.K1 and nrsD::C.K1 mutants. Mutants with the C.K1 cassette in the same orientation as the nrs genes are shown. Tenfold serial dilutions were spotted on BG11C plates, supplemented with the indicated concentrations of NiSO4, and photographed after 10 days of growth.

The amino-terminal part of NrsD is a metal binding domain. As previously mentioned, the closest NrsD homolog is the product of the A. xylosoxidans nreB gene, which has not been characterized. NrsD shows also very significant sequence similarity to several members of the major facilitator superfamily (MFS) (36). MFS transporters are single polypeptides, containing 12 to 14 transmembrane-spanning regions, capable of transporting small solutes in response to a chemiosmotic gradient. Computer analysis of the NrsD amino acid sequence indicated the existence of 12 putative transmembrane helices distributed along the first 400 amino acids of the protein (Fig. 4A). These data, taken together with the phenotype of the nrsD mutants, suggest that NrsD is a member of the MFS of permeases involved in Ni²⁺ export. Interestingly, the NrsD protein does not show sequence similarity with a family of well-characterized Ni²⁺ permeases including UreH, HupN, and HoxN (12, 16, 27). These proteins show a common topology, with eight membrane-spanning segments. The second transmembrane helix includes a putative Ni²⁺ binding motif (HX₄DH) whose mutation completely abolishes transport activity (13). This motif is not present in the putative transmembrane helices of NrsD. The strongly hydrophilic C-terminal part of NrsD contains a remarkably high number of histidine residues (12 out of 40 amino acids), which are generally considered to be potential metal ligands. In order to test whether this domain of NrsD is involved in metal binding, a chimeric protein comprising amino acids 403 to 445 of NrsD (C-NrsD) fused to the GST was expressed in E. coli (Fig. 4B). The GST-C-NrsD fusion protein was purified by affinity chromatography on glutathione-agarose. One major band of about 31 kDa (fusion protein between the GST [26 kDa] and the C-NrsD domain [5 kDa]) was visible after SDS-PAGE and Coomassie blue staining. Interaction of GST or GST-C-NrsD with Ni²⁺, Co²⁺, Zn²⁺, Cu²⁺, and Mg²⁺ was evaluated by metal affinity chromatography. For this, GST or GST-C-NrsD fusion proteins were loaded into His-bind resin chelating columns charged with the appropriate ions. About 90% of the GST-C-NrsD fusion protein was retained by the Ni²⁺-, Co²⁺-, and Cu²⁺-containing columns (Fig. 4C and D). About 60% of the GST-C-NrsD protein was also retained in the Zn²⁺-containing column. In contrast, GST-C-NrsD was not retained in the Mg²⁺-charged column (Fig. 4C and D). GST protein was not retained by any of the metal columns (Fig. 4C). These results support a role for the hydrophilic C-terminal part of NrsD as a metal binding domain. Our data suggest that this domain has a low specificity for metal binding, which has been previously shown for histidine-rich proteins (58). Histidine-rich domains have been found in UreE, HypB, and CooJ, which are small soluble proteins that are involved in processing Ni²⁺ for urease and hydrogenases (30, 42, 58). Interestingly, it has been shown that a truncated version of UreE which lacks the histidine-rich C-terminal region still binds Ni²⁺ and functions in vivo (6). Organisms with high-affinity uptake systems for Ni²⁺ have UreE-like proteins lacking the histidine-rich region (26), leading to the suggestion that the histidine-rich region functions to store Ni²⁺ ions (6). One possibility is that the histidine-rich C-terminal region of NrsD is used to store metal ions that are going to be transported.

View larger version (35K): [in this window] [in a new window]   FIG. 4.   Analysis of NrsD protein. (A) Prediction of NrsD membrane-spanning regions. The probability of transmembrane regions was calculated by using the TM-pred program (18). (B) Schematic representation of GST-C-NrsD protein. The chimeric protein comprises amino acids 403 to 445 of NrsD (C-terminal domain, C-NrsD) fused to GST. Histidine residues are underlined. (C and D) The interaction of GST or GST-C-NrsD proteins with metals was analyzed by metal chromatography. His-bind resin columns were loaded with either Mg2+, Ni2+, Zn2+, Co2+, or Cu2+. About 30 µg of purified GST-C-NrsD or GST was applied to the columns. Unbound (lanes U) (flowthrough) and bound (lanes B) (imidazole-eluted) fractions were analyzed by SDS-PAGE (12% polyacrylamide) and Coomassie blue staining (C), and protein was quantified by the method of Bradford (5). (D).

A MerR-related transcription activator involved in Co²⁺ sensing. The nrs operon was induced by Co²⁺ (Fig. 1), suggesting that it might be involved in Co²⁺ tolerance. However, nrsD::C.K1 and nrsA::C.K1 mutant cells did not show reduced tolerance for Co²⁺. These data together with the pattern of hybridization obtained with probe d in Fig. 1 suggested the existence of an alternative system involved in Co²⁺ homeostasis. The ORF slr0797 product shares clear homology with cation-transporting P-type ATPases, such as the bacterial Cd²⁺ transporter CadA or the bacterial Cu²⁺ transporters CtaA, PacS, CopA, and CopB (reviewed in reference 40). The closest homolog to the slr0797 product is ZiaA (slr0798), the Zn²⁺-dependent ATPase from Synechocystis sp. strain PCC 6803 (51), encoded by another gene of the cluster (Fig. 1A). Induction of slr0797 mRNA by Co²⁺ suggested that the slr0797 gene product might be involved in the homeostasis of this cation. This hypothesis was investigated by interrupting the slr0797 gene with a kanamycin resistance cassette (Fig. 5A) and testing the metal tolerance of the resulting mutant strain. Growth of slr0797::C.K1 mutants in Zn²⁺-, Ni²⁺-, and Co²⁺-supplemented BG11C medium was examined. Normal growth was observed in Zn²⁺- or Ni²⁺-containing medium (data not shown); however, a reduced tolerance to Co²⁺ was detected (Fig. 5B). This result indicates that the slr0797 ORF is involved in Co²⁺ tolerance. Since the slr0797 product shows clear homology with cation-transporting P-type ATPases, our data point to the slr0797 product as a Co²⁺ efflux pump. Based on this role in Co²⁺ transport, the slr0797 ORF was designed corT (for cobalt response transporter). The fact that corT is also weakly induced by Zn²⁺ suggests that this ATPase might be involved in Zn²⁺ tolerance. However, the corT mutant cells did not show reduced Zn²⁺ tolerance. A probable reason for this result is the existence of a Zn²⁺-specific ATPase, ZiaA, able to control Zn²⁺ homeostasis (51). Another possibility is that Zn²⁺ is a gratuitous inducer for corT gene expression.

View larger version (35K): [in this window] [in a new window]   FIG. 5.   Co2+ tolerance of Synechocystis corT::C.K1 and corR::C.K1 mutants. (A) Schematic representation of the corR-corT genomic region. The sites of insertion of the C.K1 cassette in the corT::C.K1 and corR::C.K1 mutants are shown. The nucleotide sequence of the corR-corT intergenic region is also shown. Putative 10 and 35 boxes of the corT promoter are boxed. A hyphenated inverted repeat (13-6-13) with one mismatch is marked with arrows. The start codons of corT and corR are underlined. (B) Co2+ tolerance of wild-type Synechocystis sp. strain PCC 6803 (WT) and the corT::C.K1 and corR::C.K1 mutants. Tenfold serial dilutions were spotted on BG11C plates, supplemented with the indicated concentrations of CoCl2, and photographed after 10 days of growth.

View larger version (63K): [in this window] [in a new window]   FIG. 6.   Sequence alignment of the CorR C-terminal domain with precorrin isomerase (CobH) amino acid sequences from different origins. COBH_SALTY, CobH from Salmonella enterica serovar Typhimurium; COBH_SYNY, CobH from Synechocystis sp. strain PCC 6803; COBH_METJA, CobH from Methanococcus jannaschii; COBH_PSEDE, CobH from Pseudomonas denitrificans. Identical amino acids are marked with asterisks; conservative changes are marked with colons or dots (as defined by CLUSTAL X [53]).

View larger version (66K): [in this window] [in a new window]   FIG. 7.   Loss of corT induction in the corR::C.K1 mutant. Total RNA was isolated from mid-log-phase wild-type Synechocystis sp. strain PCC 6803 (WT) or from Synechocystis corR::C.K1 mutant cells exposed for 1 h to a 15 µM concentration of the indicated metal ions. Control cells were not exposed to added metals (). RNA was isolated, processed, and hybridized as described for Fig. 1, using an nrsB or a corT gene probe (probes a and d, respectively [Fig. 1A]). The filters were stripped and rehybridized with an rnpB gene probe as a control.